Thursday, July 31, 2008

Cancer Data Collection Outsourced, and Then Manipulated

I have a favorite quote about the impossibly complex bureaucratic structure of the American health care system from Oxymorons by JD Kleinke,(1) (which I used in my "Cautionary Tale" article in the European Journal of Internal Medicine[2]):


Tens of thousands of well-meaning people work throughout the health care system, none of whom ever see a patient or deliver any actual medical care. They preside over an infinity of rules, regulations, forms, processes, contract outsourcing, financial brokering, benefit plan tinkering, analytical processes, incompatible data systems, and dead forests of paperwork. Health care administration in America is a tower of Babel that reaches to the moon....

It seems impossible to keep track of all the administrative, managerial and bureaucratic organizations that now impact health care. All of them, however, seem greatly susceptible to mismanagement, and worse.

Here is the latest example, from the Baltimore Sun,

A state contractor tampered with Maryland's cancer registry, a database used by researchers to track the disease's impact, counting hundreds of patients as having cancer when they did not, according to a legislative audit released yesterday.

The company, Macro International Inc., found in an internal investigation that data were deliberately altered between August 2004 and December of that year. The company fired the employee responsible for the cancer registry. State officials said that Macro employees apparently overreported the incidence of cancer to ensure that the database met standards set by a national certification association, which closely monitors registries to ensure that states have a complete count of cases.

The misinformation led researchers to send an estimated 400 women letters beginning in 2005 asking them to participate in a cervical cancer study when they did not have the disease. About 10 of those women called the state Family Health Administration, part of the state's Department of Health and Mental Hygiene - one of the first indications that the cancer registry was inaccurate.

Guy Garnett, a Macro vice president, declined to comment, citing company policy not to discuss client issues. The company completed the contract term, earning $1.9 million to manage the registry for 18 months through January. The state hired a new vendor, Westat, beginning in February.

In 2003, when Macro submitted the 2001 cancer database to the North American Association of Central Cancer Registries, it did not receive certification as required under its contract. Auditors faulted the state agency for not taking action, such as terminating the contract.

The next year, Macro did receive certification, but the auditors noted that information was later found to be erroneous. Auditors outlined several red flags that were not raised until years later.

First, the state agency did not review the 2002 data until August 2005, when it found they showed a 90 percent increase in cervical cancer cases and a 70 percent increase in melanoma cases from 1998 to 2002. At the time, auditors said, Macro could not provide an adequate explanation for the staggering rise in cancer incidence.

Around the same time, as part of the state health department's study of breast and cervical cancer diagnosis in Maryland women, about 2,300 patients in the cancer registry were asked to participate in the study. The Family Health Administration later worked with Macro to comb through medical histories, comparing original laboratory reports with data in the cancer registry, and determined that 400 of them did not have cancer. The study was suspended.

Then in May 2006, a former Macro employee informed the state agency that data were deliberately altered. Macro's investigation found that more than 13 percent of all cases in 2002 showed signs of tampering, according to the audit. Macro concluded that the changes were 'methodical and were made by one or more persons with broad access to the system,' the audit said.

I am of course familiar with the concept of a state-wide cancer registry. I must say it did not previously occur to (even) me that states would out-source the operation of such registries, and if so, that they would not bother to supervise the companies entrusted with collecting and analyzing the data.

Macro International Inc, is a for-profit, apparently privately held company. (I cannot find it on the Edgar database.) It boasts that its"mission is to deliver high-quality, research-based solutions to complex problems, integrating objective information with the advisory and implementation tasks needed to improve real world performance." That wording is so full of buzz words and management/ bureaucratic-speak that I am not sure what it means. So I cannot accuse Macro International's performance of violating its own mission. But it certainly violated the values of public health and health care research.

So we have yet another example of how health care activities have been transferred from people and organizations which were supposed to be fulfilling a (in this case, public) health mission, to (in this case, for-profit) organizations which do not have an overt commitment to the mission, a governance structure designed to fulfill the mission, or effective oversight or regulation to ensure that they fulfill the mission.

So the result was mismanagement, and fabricated data. The latter had direct impact on health care research, and on individual patients.

To take back the future of health care, we will have to ensure that health care is carried out by people and organizations pledged to uphold its core values, and working transparently and accountably under explicit and enforceable codes of ethics.

References

1. J.D. Kleinke. Oxymorons: the myth of a US health care system, Jossey-Bass, San Francisco (2001).

2. Poses RM. A cautionary tale: the dysfunction of the American health care system. Eur J Int Med 2003; 14: 123-130.